HGQ: High Granularity Quantization for Real-time Neural Networks on FPGAs — mentioned by chsun

Sat, 28 Mar 2026 16:04:40 GMT

One of the authors (of one of the two models, not this particular paper) here. Just a clarification, these models are *not* burned into silicon. They are trained with brutal QAT but are put onto fpgas. For axol1tl, the weights are burned in the sense that the weights are hard-wired in the fabric (i.e., shift-add instead of conventional read-muk-add cycle), but not on the raw silicon so the chip can be reprogrammed. Though, for projects like smartpixel or HG-Cal readout, there are similar ones targeting silicon (google something like "smartpixel cern", "HGCAL autoencoder" and you will find…

Paper and discussion history. A mention is not an endorsement.

HGQ: High Granularity Quantization for Real-time Neural Networks on FPGAs · HackerPapers discussions

HGQ: High Granularity Quantization for Real-time Neural Networks on FPGAs — mentioned by chsun